Skip to main content

2016 Press Releases

AI technology to identify child sexual abuse offenders

15 Dec 2016
Maggie Brennan, researcher and lecturer in the Schools of Applied Psychology and Criminology at UCC. Photo: Clare Keogh.

A new toolkit using artificial intelligence will help police identify new images of child sexual abuse online and find offenders who present the highest risk to the public, according to a UCC researcher behind the technology.

According to Maggie Brennan, researcher and lecturer in the Schools of Applied Psychology and Criminology at UCC, who worked on the UCC research team with Sean Hammond of UCC's School of Applied Psychology, the iCOP toolkit will automatically identify new and previously unseen images of child sexual abuse for police and help to reduce the volumes of materials specialists have to view in order to find children.

“It's common to seize computers and collections of child sexual abuse materials containing enormous volumes of illegal materials, terabytes of individual files. Having to view this material to find victims can be traumatic and distressing for the specialists working to find these children.”

Although there are already a number of tools available to help the police monitor peer-to-peer networks for child sexual abuse media, they usually rely on identifying known media. As a result, these tools are unable to assess the thousands of results they retrieve, whereas the iCOP toolkit uses artificial intelligence and machine learning to flag new and previously unknown child sexual abuse media.

The new approach combines automatic filename and media analysis techniques in an intelligent filtering module. The software can identify new criminal media and distinguish it from other media being shared, such as adult pornography.

According to Brennan, “law enforcement urgently need these kinds of supports to help them manage the volumes of cases they are being faced with - to find the children who are victimised in these images and videos, as well as those offenders who present the highest risk to the public.”

The research behind this technology was conducted in the international research project iCOP – Identifying and Catching Originators in P2P Networks – founded by the European Commission Safer Internet Program by researchers at UCC, Lancaster University and the German Research Center for Artificial Intelligence (DFKI).

The team at UCC worked closely with international law enforcement specialists in online child sexual abuse investigation, to understand their needs and develop a tool that allows them to find the most urgent cases for intervention. “Our role also involved developing a psychological profiling system to identify viewers of child sexual abuse images who may be at risk of committing hands-on abuse.”

“We have been researching this topic with international law enforcement agencies like Interpol for many years, since the early 2000's. The volumes of child sexual abuse images and videos now in circulation is a real concern, and it can be overwhelming for law enforcement. Trying to find recent or ongoing cases of child sexual abuse is an absolute priority, but the sheer volume of illegal materials in circulation online makes this task incredibly difficult for the police,” Brennan said.

There are hundreds of searches for child abuse images every second worldwide, resulting in hundreds of thousands of child sexual abuse images and videos being shared every year. The people who produce child sexual abuse media are often abusers themselves – the US National Center for Missing and Exploited Children found that 16% of people who possess such media had directly and physically abused children.

“Identifying new child sexual abuse media is critical because it can indicate recent or ongoing child abuse,” explained Claudia Peersman, lead author of the study from Lancaster University. “And because originators of such media can be hands-on abusers, their early detection and apprehension can safeguard their victims from further abuse.”

The researchers tested iCOP on real-life cases and police trialed the toolkit. It was highly accurate, with an error rate of only 7.9% for images and 4.3% for videos. As the system reveals who is sharing known child sexual abuse media, and shows other files shared by those people, it will be highly relevant and useful to police.

The toolkit is described in a paper published in Digital Investigation.

For media requests, contact Lynne Nolan, Media & PR Officer, UCC on 087-210 1119.

University College Cork

Coláiste na hOllscoile Corcaigh

College Road, Cork T12 K8AF